Virtual assistants with artificial intelligence integrate quickly and easily to the most used messaging channels in the market (WhatsApp, Webchat, Facebook Messenger, private Twitter, among others); internal systems, payment gateways and more, thanks to the conversation flow design tool.
It designs conversational experiences that allow you to improve and automate the interaction with your customers, combining chatbots and human agents.
Conversational flows are implemented based on the client’s needs. Virtual assistants understand the written word through understanding processes.
S1 virtual assistants can be designed to work on open as well as directed conversational flows.
For certain themes it is advisable to propose an open language design to achieve greater naturalness in the dialogue. In topics associated with specific processes, it is convenient to guide the user in a flow directed by menus, to reach their objective faster.
IA Conversational
in 5 steps
PHRASE STANDARDIZATION
NLP
01
.
For natural language understanding, S1 normalizes the requests of incoming users.
ATTEMPT DETECTION PROCESS
NLP
02
.
Uses NLP to detect objects in the sentence (nouns, verbs, adjectives, etc.). NLP returns objects included in the sentence.
PHRASE
MAPPING
03
.
With rthe target discovered, the engines searches the conversational model for a next match on a preloaded phrase.
SIMILARITY
04
.
If the phrase does not match exactly, the engine searches the model for the most similar phrase.
MACHINE
LEARNING S1
05
.
In case it is not detected by similarity, Machine Learning is used with configured hit thresholds to determine the problem expressed by the user.
Normalization process
Allows you to identify specific situations during the conversation with the user in order to take specific actions during the course of the conversation if, for example, the process already identifies that we are dealing with a purchase request, complaint or inquiry.
- Forced spelling
- Empty words
- Welcome words
- Thank you words
- Request words
- Action words
- Farewell words
- Insults
- Synonyms
NLP + Object Detection
Natural Language Processing aims to optimize the communication between the human interlocutor and the artificial intelligence by means of a natural language (Spanish, English, etc.).
To achieve this, S1 Bot reduces to the minimum expression what the user wants, in order to achieve an effective communication with the AI.
After going through the NLP process, the S1 engine goes through the following steps:
phase mapping, similarity, Machine Learning.
Sentence mapping
and similarity
It consists of a logical comparison of the input stimulus with a set of rules that, if verified as true, will determine which intention resolves the specific stimulus. These conditions may involve “Same as” (phrase match) or “Similar to” (similar phrase).
Machine
Learning
S1 has a patented machine learning approach to intent detection, which consists of a broader search in the conversational model developed for the client, where our engine analyzes all preloaded phrases and intents to find the most likely intent and its assertiveness rating.
With this approach, each specialized bot can access a highly focused set of data that boosts the assertiveness rate.
Boost customer experience and reduce costs with S1 Bot
Design a conversation and deploy it across multiple channels
S1 Bot improves the customer experience across all conversational channels, no matter how your customers communicate.
We automate and streamline
Reduce average handling time per contact by up to 40% by automating case management and resolving repetitive tasks.
Streamline conversations with your customers, helping them understand their issues and resolving them every step of the way.
S1Bot empowers agents
It is intelligent, knowing when to intervene and manage the conversation, and when to refer it to a live agent.
Tailored integrations with all systems
The bot integrates with your backend to execute repetitive tasks and help customers with their queries.
Secure transactionality
S1 operates in a PCI-DSS certified environment, protecting sensitive data and operating securely.
Design conversational care experiences
The conversational flow designer enables self-management for configuration, administration and training of S1 Bot.
Design conversational care experiences.
With corresponding trainings, the solution provides the ability to create, modify and train one or multiple bots to work alongside the agent team.
+10 million incoming messages per month
+60 simultaneous messages handled per minute
+86% average understanding
